Directory Structure

.
└── Search-based Planning
    └── Search_2D
        ├── bfs.py                                  # breadth-first searching
        ├── dfs.py                                  # depth-first searching
        ├── dijkstra.py                             # dijkstra's
        ├── a_star.py                               # A*
        ├── bidirectional_a_star.py                 # Bidirectional A*
        ├── ARAstar.py                              # Anytime Reparing A*
        ├── IDAstar.py                              # Iteratively Deepening A*
        ├── LRTAstar.py                             # Learning Real-time A*
        ├── RTAAstar.py                             # Real-time Adaptive A*
        ├── LPAstar.py                              # Lifelong Planning A*
        ├── D_star.py                               # D* (Dynamic A*)
        └── D_star_Lite.py                          # D* Lite
    └── Search_3D
        ├── Astar3D.py                              # A*_3D
        ├── bidirectional_Astar3D.py                # Bidirectional A*_3D
        ├── RTA_Astar3D.py                          # Real-time Adaptive A*_3D
        └── LRT_Astar3D.py                          # Learning Real-time A*_3D
└── Sampling-based Planning
    └── rrt_2D
        ├── rrt.py                                  # rrt : goal-biased rrt
        └── rrt_star.py
    └── rrt_3D
        ├── rrt3D.py                                # rrt3D : goal-biased rrt3D
        └── rrtstar3D.py
└── Stochastic Shortest Path
    ├── value_iteration.py                          # value iteration
    ├── policy_iteration.py                         # policy iteration
    ├── Q-value_iteration.py                        # Q-value iteration
    └── Q-policy_iteration.py                       # Q-policy iteration
└── Model-free Control
    ├── Sarsa.py                                    # SARSA : on-policy TD control
    └── Q-learning.py                               # Q-learning : off-policy TD control

Animations

DFS & BFS (Dijkstra)

• Blue: starting state
• Green: goal state

A* and A* Variants

Value/Policy/Q-value/Q-policy Iteration

• Brown: losing states

SARSA(on-policy) & Q-learning(off-policy)

• Brown: losing states

Papers

• Anytime Repairing A*: by Maxim Likhachev, Geoff Gordon and Sebastian Thrun
• Lifelong Planning A* by Sven Koenig, Maxim Likhachev and David Furcy
• D*: by Anthony Stentz and Is Carnegle Mellon
• Focussed D*: by Anthony Stentz
• D* Lite: by Sven Koenig and Maxim Likhachev
• Field D*: by Dave Ferguson and Anthony Stentz
• Anytime D*: by Maxim Likhachev, Dave Ferguson, Geoff Gordon, Anthony Stentz and Sebastian Thrun
• Theta* & AP Theta*: by Alex Nash, Kenny Daniel and Sven Koenig
• Lazy Theta*: by Alex Nash, Sven Koenig and Craig Tovey
• Incremental Phi*: by Alex Nash, Sven Koenig and Maxim Likhachev

License

MIT License